LHC Computing Grid Project

Project Execution Board

Notes of the meeting of Tuesday March 16, 2004

DRAFT 5 22/MAR/2004

 

Present:

Dario Barberis, Ian Bird, Philippe Charpentier, Frédéric Hemmer, Bob Jones, Jürgen Knobloch (secretary), Massimo Lamanna, Alberto Masoni, Mirco Mazzucato, Bernd Panzer, Les Robertson (chair)

 

Actions: Actions are identified by bold blue italics.

Minutes of last meeting and matters arising

The minutes of the meeting of the 2nd March were accepted.

 

Major decisions from recent meetings

GDB – Mirco

Mirco summarized the main points of the GDB meeting of March 8.  (document).

There were many reactions from the experiments on the security document "LCG User Registration and VO Management". The experiments want to have a centralized system for the LHC experiments as much as possible connected to the CERN registration (“Human Resources”) database which contains already the affiliation of people to experiments. The registration and the VO attribution should be handled in one place as outlined in Appendix A of the document.

A questionnaire concerning requirements for IP connectivity to the worker nodes has been sent out – replies from some experiments are still outstanding. Ian would like to have the complete picture with the requirements of the experiments on one side and the constraints of the computing centres on the other.

With the grid deployment in EGEE, a policy body like the GDB will be required – maybe an extended common GDB. Les, Ian, Mirco and Dave Kelsey will discuss with the EGEE management.

Concerning the LCG-2 deployment, a mismatch in the queuing systems at some centres has been experienced. For the Alice data challenge this is currently circumvented by manual intervention.

POB - Les

The summary of the POB meeting of 17-18 February is summarized in a document. Concerning point 4 of the document, it is decided that the LCG quarterly reports should from now on contain the regular summary reports of the plans of experiments for integrating Applications Area products in their applications.

Level-1 milestones

Jürgen has updated the milestone table taking into account the input from PEB members. This final table was approved by the PEB to be sent to the LHCC referees. The referees will be informed that more concrete milestones for the Applications Area will be supplied as soon as the plan is ready.

Bob announced that at the recent NA4 meeting, it was agreed that  "The major deliverable will be the evaluation report, due Jan 2005, reporting on the applications use of LCG/EGEE. This will have to be based largely on data challenge work in 2004 with current middleware, and supplemented by first work done with new EGEE middleware." Therefore the experiments should agree to make material (grid usage results and feedback) on the use of the LCG-2 for their data challenges in 2004 available to the NA4 people for inclusion in this deliverable.

Philippe reiterated the LHCb concern about the heavy manpower requirements to satisfy simultaneously the parallel routes of milestones 2 and 3.

 

Tier 1 services for Tier 2 centres - Les

Recently, it became clear that the responsibilities for supporting Tier-2 centres were not clearly defined. Les has drafted a first document collecting items to be addressed by a small group of four experts from Tier-2 an Tier-1 centres. The resulting document should also be used as background information by the MoU task-force because the required support will need to be properly funded.

Following the discussion in the PEB and with further input by mail, Les will provide an updated version of his proposal and contact people to join the task force.

Point raised include:

  • consider not only the steady state but also the start-up phase of a Tier-2,
  • in large countries having only Tier-2 centres (like Russia) one of them would take up the supporting role – so one would rather call the supporting sites “primary sites” instead of “Tier-1s”
  • the question is also relevant for EGEE
  • does this also include software support and training?
  • what about application support (role of experiments)?

Status of data challenges

Ian said that by now 1800 CPUs are running LCG-2 at the core sites. Some batch queue configuration questions are being investigated.

Alberto said that Alice had in the last days more than 1000 concurrent jobs running using 1.5 THz of CPU power (1000 processors of 1.5 GHz). A number of problems have been found and fixed. One centre advertising too much capacity can prevent others to receive jobs – this is currently prevented by manual intervention. A proper solution is being investigated. The Alice data challenge is currently stopped for a day to solve a file transfer issue.  Alberto said that, keeping the present production rate, under stable conditions, the data production could be finished in about one month

On the question when Karlsruhe would move to LCG-2, Alberto replied that once some outstanding problems at CNAF and CERN are fixed, Alice will also use LCG-2 at Karlsruhe.  

Ian mentioned that the large size of Alice jobs (sometimes over a GB) required a change in the queue configuration.

Dario said that ATLAS is now preparing for the data challenge starting beginning of May.

LHCb wants to start testing very soon. They plan to provide next week a release of the application software for the data challenges.  On the question of Philippe concerning mass storage space, Ian said that the plan is to move to the SRM interface to MSS by the end of March – also at RAL.

A.o.b.

Massimo has finished a first round of discussions with the experiments on Arda. He appreciated their constructive attitude. He expects to present a plan next week.  An agreement has been reached on how people are assigned to experiments. The office space for the project is still to be solved.

 

Actions

#

Date opened

Description

Responsible

Date closed

1

16dec03

ALICE, CMS and LHCb to name someone responsible for coordinating deployment on LCG-2

Federico, David S., Philippe

Done

2

16dec03

Understand why the substantial resources in Liverpool are not available for LCG-2.

6jan04- visit to RAL organised for 24jan04

Les

Done

3

16dec03

Confirm that the absence of BNL in the LCG-2 deployment list is due to manpower shortage

Les

Done

4

16dec03

Experiments to request through their national contacts that their resources in the core LCG-2 centres are integrated in LCG-2

Federico, Dario, David S., Philippe

Done

5

16dec03

Regional centres to be asked to clarify their mass storage plans.

Presented by RCs in GDB of 13jan04

Les

13jan04

6

12jan04

Revised proposed GAG mandate

Federico

27jan0

7

27jan04

Revised ARDA note

Les

12feb04

8

27jan04

Establish a weekly “Deployment Meeting”

Ian

2feb04

9

27jan04

Note on new project proposal from Trento

Federico

 

10

12feb04

Define new name for middleware

Bob

 

11

12feb04

Nominate Arda contact persons

Alice

 

12

12feb04

Nominate people for Phase 2 requirements of the experiments

Experiments